Multiple sclerosis (MS) is an inflammatory disease of the CNS. It is thought that tissue injury occurs when activated, myelin-reactive T cells migrate into the CNS and cause damage to myelin, oligodendrocytes and axons. Ultimately, MS is a complex genetic disease as studies in twins, half-siblings, and adoptees indicate a strong family inheritability. However, large-scale studies attempting to identify genes affecting the disease have so far had limited success, calling for a more powerful search strategy. The classic method of finding genes--linkage mapping--works well for rare, single gene disorders that run simply in families. But linkage scans have failed to find the genes for more common, genetically complex diseases including MS. The approach most likely to work for gene discovery is the direct assessment of variation in populations and its association to disease. With present technology, the best-known way of doing this---haplotype mapping--is not practical because it requires studying too many sites in the genome. Because MS is significantly more common in Europeans than in Africans, a new approach, admixture mapping, may be a shortcut for using association studies to find disease genes. Specifically, we hypothesize that the intermediate genetic risk of MS in African Americans is derived almost entirely from their small percentage (10-40%) of European ancestry. By scanning along the genomes of African Americans with MS looking for regions of unusually high European ancestry, we can identify the 'European' gene segments that are likely to contain the genes that are related to MS risk. In this study, we propose to carry out the first whole-genome admixture scan for human disease genes, using 100-times fewer markers than a haplotype-based study. An admixture scan has the potential to rapidly identify disease regions especially for the subset of diseases that have different prevalences in two populations. The admixture mapping approach has only become feasible in the past year because of the large numbers of SNPs discovered with known frequencies in both African- and European-Americans. The SNP resources and novel analytical tools have now converged with large sample collections of African-American MS patients. The central aim of this project will be to carry out an admixture scan for MS genes in a sample of 1,000 African Americans with MS and 340well-matched controls. To follow-up all the genomic regions associated with disease, we will triple density of markers to increase statistical confidence in the results and refine the positions. We will then move to a targeted haplotype-based association study in the most interesting regions to clone new genes associated with MS.